Last-word prediction on LAMBADA es
16Accuracy (Zero-Shot)MGPT-2
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| MGPT-2#Params=124M, Patch size=-, zero-shot evaluation=true2026.04 | 16 | — | — | |
| GPT-2#Params=124M, Patch size=-, zero-shot evaluation=true2026.04 | 6.1 | — | — | |
| MIXAR stage2#Params=477M, Patch size=32×32, zero-shot evaluation=true2026.04 | 4.1 | 55.4 | 58.9 | |
| MIXAR stage2#Params=116M, Patch size=32×32, zero-shot evaluation=true2026.04 | 2.8 | 63.5 | 68.1 | |
| PIXAR stage2#Params=113M, Patch size=8×8, zero-shot evaluation=true2026.04 | 2.5 | 54.7 | 58.8 | |
| MIXAR stage1#Params=477M, Patch size=32×32, zero-shot evaluation=true2026.04 | 1.2 | 48.8 | 52.6 | |
| MIXAR stage1#Params=116M, Patch size=32×32, zero-shot evaluation=true2026.04 | 0.5 | 46.5 | 50.2 |